Acute lymphoblastic leukemia (ALL) is the most common malignancy among children and young adolescents. More than 90% of children are cured with a combination of multiple chemotherapeutic drugs. Some may suffer from debilitating toxicities due to the cytotoxic drugs, necessitating pharmacogenetic research to search for genetic markers predictive of treatment toxicities, so that treatment can be better tailored based on patients? genetic makeup. Bone toxicities, including osteonecrosis and fractures, most often due to glucocorticoids, are the most common complications in ALL, which have long-lasting detrimental impact on the still developing skeletons in children. In DFCI ALL Consortium Protocol 05-001, a multi-center clinical trial for childhood ALL conducted in Canada and the US, we found that Hispanic children had less drug toxicities to their bones than non-Hispanic children. Although reasons for the observed ethnic differences are largely unknown, patients? inherited genetic background may be at play. This R03 grant seeks to perform a novel pharmacogenomic study based on 05-001 and its successor trial DFCI 11-001 to identify genetic underpinnings of ethnic disparities in bone toxicities. We plan to first test global genetic ancestry with bone toxicities, followed by a bivariate genome-wide association study (GWAS) to jointly analyze osteonecrosis and fracture as two related traits. This approach has been proven to be highly effective in ascertainment of genetic causes of ethnic disparities and further identification of related genes and loci. Specifically, we propose the following four Aims. Aim 1. Determine whether the composition of genetic ancestry in Hispanic children is an underlying cause for the ethnic disparities in therapy-related bone toxicities, namely osteonecrosis and fracture, in children with ALL in the DFCI 05-001 and 11-001 trials. Aim 2. a). Investigate single variants and polygenic risk scores from previous GWAS of bone-related phenotypes with bone toxicities in children treated for ALL; b). Identify genetic loci associated with therapy-related bone toxicities by performing bivariate GWAS analyses with directional alignment and meta-analysis in the DFCI 05-001 and 11-001 trials; c). Investigate whether genetic variants and polygenic scores significant in 2a and 2b explain disparities in bone toxicities between Hispanic and non- Hispanic children with ALL. Pharmacogenetic and pharmacogenomic research in childhood ALL has been remarkably successful and highly rewarding in identifying risk variants with large effect size for treatment outcomes and toxicities based on patient populations of relatively small sample size. This is likely due to high susceptibility of children to cytotoxic effects of chemotherapeutic drugs and thus much larger effect size each variant carries. Our study will be the first to focus on ethnic disparities in bone toxicities for children treated for ALL. We expect to elucidate genetic mechanisms underlying the ethnic disparities in therapy-related bone toxicities between Hispanic and non-Hispanic children with ALL. The identified risk variants may have translational significance to guide the treatment for children with ALL.